Quality-sorted Stacking is Better

19 Mar 2014.  I’ve been playing with the image of M3 I took last week, trying to improve on the stacking.  By way of review, this was 300 30-sec subframes, and originally I stacked them in Nebulosity 3.2 and used all of them.  I knew this included some bad subframes, and just figured that standard deviation clipping would get rid of the bad data.  While I’m sure that did help, I decided to do a little filtering.  I like the quality sorting algorithms in Deep Sky Stacker, and although I have not used DSS much lately, I decided to give it a try (it is a lot less painful to use VMWare/Windows 7 on my Macbook now that I have a solid state drive).  I got a much sharper image from DSS.  I saved it as a 32-bit/channel rational FITS file out of DSS, then processed as usual in StarTools.  I’m having trouble getting as much color to show up as with the Nebulosity-stacked image, but looking at a bunch of M3 images on Astrobin I think the DSS-stacked version is closer to that average.  Here they are together so you can decide:

The Nebulosity-stacked image (above) is certainly more colorful, but the more selective DSS-stacked image has sharper, better-resolved stars.  I wish I could get Nebulosity’s colors with DSS’ sharpness.  I guess I could use DSS’ quality sorting to decide which subs to stack in Nebulosity, then have the best of both  worlds.  It’s a bit of a pain, but I could try that.  

Here it is:

It's the best of the lot, in my opinion.  I did encounter what appears to be a bug in Nebulosity 3.2: when I checked off subframes to skip and then tried to use the automatic alignment star selection, the selected star was lost when the routine came to the skipped subframes.  I guess I should report this.